25 April 2023
Vienna
Austria

Data Science Day 2023-1

In-person learning and networking experience for data practitioners in the news

Programme

 

Data Science Day 2023-1

April 25 Tuesday

09:30

Registration and welcome

ARCOTEL Wimberger Neubaugürtel 34-361070 Vienna (room Wimbergersaal)

10:00

Beyond Quality Read: How GEDi is developing attention metrics to evaluate readers' engagement

Wimbergersaal

Engagement scores are essential metrics to understand how users interact with websites, providing a proxy of interest to subscribe; page views, unique users and avg time of session are the most common metrics adopted to measure engagement, not taking into account qualitative aspects of reader interactions. The fundamental question to assess user engagement properly should be “how does a user read?” rather than “how much does a user read?”. This shift implies that we must measure the intensity of interactions and the quality of the page view, mixing information collected from readers’ actions. This approach is the starting point of GEDI’s journey, from the concept to the first release, to build a quality engagement metric to support newsroom and marketing in the daily routine.

Speakers
Data Analytics Manager | Data Science Lead at Gedi, Italy
10:30

Mediahuis' personalization algorithm

Wimbergersaal

Speakers
Head of Data Science & AI, Mediahuis, Belgium
11:00

coffee break

11:15

Get-to-Know breakouts

Wimbergersaal

12:00

NLP at Amedia

Wimbergersaal

Amedia is a Norwegian conglomerate of over 100 local newspapers. In his presentation, Emiliano Guevara will give a run down of Amedia’s current NLP stack, including data collection, model training, and deployment in our production systems. Some of Amedia’s key applications are document classification, modelling reader interests, and predicting popularity and sales.
Speakers
Senior Data Scientist at Amedia, Norway
12:30

Large Language Models: an overview on evolutions and what to expect

Wimbergersaal

An overview of the evolution of language modelling, how transformers work and why they work so well, as well an overview of what LLMs can do and use cases of how you can use them

Speakers
Machine Learning Engineer at Google, Germany
13:00

Buffet Lunch & Networking

Wimbergersaal

15:00

Building a collaborative data culture between data and newsroom

Wimbergersaal

Integrated data analytics application in the editorial system. CNC’s Carl’s app integrates all data into one place, personalized for authors, editors, and editors-in-chief

Speakers
Senior Digital Content Analyst at Carl, Czech Republic.
16:00

First round of topical breakout roundtables

Wimbergersaal

Participants will join pre-selected roundtable breakouts to discuss and learn from each other’s experiences:

  1. Communicating with the newsroom: dash boarding, real-time restitution of sales and traffic data. Visualization and storytelling for the newsroom.
  2. Defining meaningful and purposeful business KPIs. Which models do you use to forecast subscriptions and revenues strategically?
  3. Natural Language Generation. What kind of ChatGPT use cases do you know in analytics and data, and which use cases do you discuss with other teams, editors or sales?
  4. Subscription, propensity, predicitive models, intelligent use of data for targetted acquisition and churn reduction measures.
16:45

Break

17:00

Second round of topical breakout roundtables

Wimbergersaal

Our last round of topical roundtable discussions will be based on the day’s discussions. We have the following preliminary themes for you, but we will focus on the topics you may have discussed earlier in the day.

  • Customer target acquisition: Which models and use-cases are currently implemented?
  • Google Analytics, GA4.
  • Topic Modeling: Which use-cases are currently implemented and in use?
  • What is your experience with Marketing Mix Models/Media Mix Models? How do you deal with online vs. offline advertising channels?
  • AutoML: Which methods are used to speed up the optimization process and improve the performance of models?
  • How and where do you consolidate different customer IDs? How is your experience with a Customer Data Platform (CDP)?
  • Align content more strongly with the user and intent using AI (different targeting/playout depending on search intent – tools such as ChatGPT).
  • Ready-made solutions vs in-house-made solutions?
  • ML models for text understanding.
  • ….
18:00

Wrap-up of the day

18:30

Networking & drinks sponsored by Chartbeat

Join us for drinks and network with old and new friends!